Introduction
From 1890 to 1921 immigration was open, and more than 17 million people arrived in America (Goldin Reference Goldin, Goldin and Gary1994: 223). Debarment and deportation rates were under 3 percent (Table 1). The Bureau of Immigration was nascent, established in 1891 by Congress in the Treasury Department to take enforcement from the states (Ettinger Reference Ettinger2009: 68). In 1906, this drastically changed when Congress assumed authority over naturalization and combined both functions in the new Bureau of Immigration & Naturalization (BIN), part of the Department of Commerce & Labor (US Congress 1906c). BIN produced a higher denial rate for naturalization, over 20 percent (Fig. 4), responding to Republican concerns about fraud in the courts and new citizens who could tip the scales in elections, jeopardizing party control of Capitol Hill and the White House (House Committee on Immigration and Naturalization 1906a: 99; “Report to the President” 1905: 11–15).
This narrative focuses on BIN, from 1907 to 1913, when it was disbanded. My core argument is that Republican officeholders used bureaucratic reorganization to nimbly influence the implementation of immigration and naturalization policies. This better explains the range of enforcement activities. My examination of the Bureau is important for the literature on immigration history because this scholarship concentrates more on legislative, than administrative, means to explain policymaking (Feys Reference Feys2013: 9–10; Lee Reference Lee and Ueda2006: 7–8; Marinari Reference Marinari2020: 15). Given this difference, I rely on political control literature, which considers congressional and presidential power over bureaucracy. This includes the use of instruments like structure (McCubbins et al. Reference McCubbins, Noll and Weingast1989). Since it has been underexplored (Moe Reference Moe2012), this research benefits both literatures.
To probe how BIN was used, I supplemented secondary sources with archival records from the Roosevelt and Taft presidencies, from Congress, and from the Bureau. These were selected because the focus is on political control of bureaucracy. The records were among the presidents, Congress, political appointees, Bureau staff, and lobby groups, and included correspondence, legislation, congressional hearings, regulations, memoranda, and reports. Careful attention was paid to documents between appointees and staff since bureaucratic autonomy can be masked by political harmony between bureaus and politicians (Carpenter Reference Carpenter2001: 357). Because official documents could have been misfiled, lost, or destroyed, I reviewed collections of personal papers of key political appointees: secretaries of Commerce & Labor Oscar Straus and Charles Nagel, Ellis Island commissioners Robert Watchorn and William Williams, and Information Division Chief Terence Powderly.
My examination of these sources is unique because I apply an analytical framework, “policy innovation through bureaucratic reorganization,” which I developed. The framework’s evolution phase, consisting of three steps, will be discussed: (1) resource adjustment, (2) coalition management, and (3) system redesign (Hernandez Reference Hernandez2015: 16). While they follow a linear progression, there is a feedback loop between (1) and (2). Agency resources are altered in response to lobby pressures, and this can be repeated. Then (3) results, except this step is refined herein to fully consider an agency’s role when it is restructured.
This study reveals that presidents Theodore Roosevelt and William Howard Taft used BIN’s structure to pivot between easing and tightening immigration and naturalization enforcement. Since 1907, this was facilitated via centralization of these functions in the Bureau and modifications to its budgets, staffing, and infrastructure (resource adjustment). Still, expansion and reduction of these proved imperfect to similarly increase and decrease enforcement outputs. Nonetheless, such outputs furthered presidential efforts to placate interest groups that favored or disapproved of immigration (coalition management). In 1913, immigration restrictionists, led by the American Federation of Labor, secured decentralization of these functions from Congress (system redesign). BIN contributed to this by producing adverse aftereffects on immigrant communities.
Next, I will explain the insights I rely on from both literatures to support the framework, and how the framework can add to these literatures.
Theoretical framework
To appreciate exactly how the state matters (Skocpol Reference Skocpol2008 [1985]), the literatures on political control and immigration history are combined to probe immigration bureaucracy. To assess political control, my framework’s resource adjustment step focuses on structure: “The allocation of resources and decisional authority” in an agency (McCubbins et al. Reference McCubbins, Noll and Weingast1989: 431–32). While the effect of structure on agencies “created long ago” can be questioned (Carpenter Reference Carpenter2001: 357–59), the study of new bureaus is promising. To analyze a bureau’s influence on lobby pressures, my coalition management step considers structure, too. Interest groups play a role in reshaping bureaucracy (Moe Reference Moe, John and Paul1989), and their activities are formed by its structural arrangement (Macey Reference Macey1992). To better understand reorganization, my system redesign step reviews its development not only from outside, but also inside, a bureau. This is facilitated by scrutinizing operational tweaks in an agency since they can spur institutional change (Mahoney and Thelen Reference Mahoney, Thelen, Mahoney and Thelen2009).
For resource adjustment, the starting point of analysis is the location of decisional authority in the bureaucracy. Presidents seek centralized control for unitary action, while Congress wants “particularized” power for impromptu interventions (Moe Reference Moe, John and Paul1989: 277–81). Centralization “economize[s] on the transmission and handling of information” (Arrow Reference Arrow1974: 69), which can improve intra-agency coordination (Nou Reference Nou2015: 451). The president uses centralized decision-making to shift gears between accelerating and decelerating enforcement, and Congress counters by limiting bureaucratic discretion (Whitford Reference Whitford2005: 30, 44–45). These shifts show how contested political control is; yet, understanding their impact on service coordination is equally important.
Next, the analysis focuses on the allocation of budgets, staffing, and infrastructure. Budgeting affects resource availability for implementation. Funding reductions signal that enforcement activities should be curtailed while funding increases are not as effective in expanding enforcement (Wood and Waterman Reference Wood and Waterman1993: 512). Such signaling is further limited: the processing of information is slowed in agencies because of hierarchical layers, delaying responses to budgetary changes (Carpenter Reference Carpenter1996: 284, 299). Such funding changes are examples of “active” control since politicians use them to induce agency compliance with their policy preferences (Calvert et al. Reference Calvert, McCubbins and Weingast1989: 595, 602, 604–5). Bureaus are susceptible to budget control mechanisms (Moe Reference Moe2012: 4–6), even when agency officials seek to maximize their budgets and presumably have leverage over Congress (Niskanen Reference Niskanen2008 [1971]). However these mechanisms may not be as successful for interagency work because of conflicting goals.
Staffing includes political appointees and careerists, and they can be at odds (Moe Reference Moe, John and Paul1989: 282–84). Because this tension compromises political control, it requires more scrutiny. Presidents politicize bureaucracy by expanding the number of appointees and their managerial responsibilities. They “interpret the vague and sometimes conflicting laws” and “change public policy administratively.” But careerists have expertise, including “a better understanding of agency programs.” If they are minimized by politicization, when Congress largely confirms presidential appointments, then agency performance suffers (Lewis Reference Lewis2008: 2, 7, 203–5). This can also occur when mid-level and front-line careerists develop autonomy from appointees (Carpenter Reference Carpenter2001) and workarounds from protocols (Lipsky Reference Lipsky2010 [1980]: 83), respectively.
Infrastructure is underexamined in political control literature. My resource adjustment step regards infrastructure as facilities where employees work. Such places are part of the national infrastructure; Alexander Hamilton saw its development in the public interest (Light Reference Light2008: 7). Akin to my conceptualization, this physical infrastructure has a human dimension since employees help to run the government (Verkuil Reference Verkuil2017: 6). This governmental infrastructure includes “agencies, people, and processes” (Lewis Reference Lewis2019: 783). These ideas help to contextualize infrastructure in and to consider their effects on, immigration bureaucracy. This is crucial since organizational structure is not neutral, and “will bias policy-making toward some outcomes and away from others” (Hammond and Thomas Reference Hammond and Thomas1989: 158).
For coalition management, the analysis centers on the interplay between interest groups and bureaucracy. It is useful to see both, respectively, on the demand, and supply, side. Lobbies press politicians for action and, like “a market,” they respond with “products” (Weingast Reference Weingast1981: 161). Immigration policymaking is characterized by client politics; interest groups that favor inflows dominate and receive concentrated benefits, while the public bears concentrated costs. This contributes to liberal immigration policies (Freeman Reference Freeman1995: 885–86). When immigration and naturalization enforcement outputs are considered benefits, and their numbers shift, this may signify a change in the mode of immigration politics. This can be explained by the bureaucratic arrangements interest groups are in. When regulatory agencies oversee single industries, political control is best since agencies designed for multiple industries are more susceptible to lobbies (Macey Reference Macey1992: 93–94). Structural decisions merit close attention; they show how lobbies protect their policy preferences or expose their opponents’ choices (Moe Reference Moe, John and Paul1989: 273–77).
For system redesign, the analysis gives equal attention to internal and external forces. I supplement my framework with institutional change scholarship, which appreciates that incremental developments in bureaus can produce “fundamental transformations” (Mahoney and Thelen Reference Mahoney, Thelen, Mahoney and Thelen2009: 2–3). This happens through the process of “conversion” since policies are redirected for new ends. Otherwise, this can be missed because contestation is less visible in administrative, than in legislative, arenas (Hacker et al. Reference Hacker, Pierson, Thelen, Mahoney and Thelen2015: 185–86). The political control and immigration history literature emphasize institutional change via external forces. In one view, structural reforms of public agencies are a perpetual process, driven by interest groups and decided by officeholders (Moe Reference Moe, John and Paul1989: 284–95). In another, institutions are largely stable because of interrelated processes, including structural opportunities and constraints and international threats. They produce reforms, too (Tichenor Reference Tichenor2002: 294).
In the three sections that follow, I will use my framework’s steps to analyze how BIN pivoted, why it did, and what the consequences were for newcomers and the agency.
Resource adjustment
Presidents Roosevelt and Taft faced several immigration enforcement challenges, especially inflows on steamships, migrant border-crossings, and a new naturalization process (Ettinger Reference Ettinger2009: 68–73; Schneider Reference Schneider2001: 50–58). In this section, I will explain BIN’s enforcement pivots through its interconnected parts of centralization, staffing, budgeting, and infrastructure (resource adjustment). The consolidation of immigration and naturalization enforcement and the presidents’ political appointees mostly proved successful in shifting enforcement activities. However, the infusion of front-line staff and funding did not consistently result in increased enforcement. It was stymied by a slow-evolving infrastructure, too.
Centralization of immigration and naturalization functions facilitated enforcement pivots, but not coordination of these functions. Appointees readily directed these pivots because centralized decision-making expedites information processing in hierarchical organizations (Arrow Reference Arrow1974). This is distinct from specialization: the process of allocating tasks across an agency for optimal coordination (Nou Reference Nou2015: 459). BIN’s architect, Representative William Bennet (R-NY), aimed for these functions to reinforce each other. To safeguard naturalization, the investigation and processing of immigrants who applied for citizenship was enabled because their arrival records were available in one agency (House Committee on Immigration and Naturalization 1906a: 33, 41–42). Since naturalization records were there, too, these would help expose newcomers who fraudulently claimed to be US citizens to gain entry (Feys Reference Feys2013: 302). As will be discussed in two subsections that follow, the Bureau did not achieve such coordination despite Congress gifting these presidents centralized power.
Easing and tightening immigration enforcement
Between 1892 and 1906 rates for debarment – the exclusion of arrivals – and deportation – expulsion to their countries of origin – were mostly under one and 0.10 percent (Table 1).Footnote 1 Debarment was weakened because business interests commercialized remote control, the government’s “restrictive capacity” abroad (Schneider Reference Schneider2011: 10–11; Zolberg Reference Zolberg2006: 200). Deportation was minimized; in 1903, the statute of limitations was extended from one to three years (Moloney Reference Moloney2012: 243). Still, between 1907 and 1913 these rates rose up to 2.48 and 0.12 percent, respectively (Table 1); debarment and deportation numbers were lower between 1907 and 1909 (“first period”) compared to 1910 through 1913 (“second period”) (Figs. 1–2).Footnote 2 Roosevelt and Taft’s appointees directed these functions to focus on “new” immigrants from Eastern and Southern Europe (Zeidel Reference Zeidel2004: 6–7), perceived as poor and infirm. Except Roosevelt’s team was more liberal with enforcement. Taft’s was strict and spurred on when Congress eliminated BIN’s budgetary discretion, which disincentivized admission of newcomers for head tax revenue. Also, Ellis Island’s medical operations were upgraded and put to full use.
Figure 1. Debarment numbers and reasons (left scale) and admissions (right scale).
Source: Own elaboration based on ARCGI, 1906–14.
Political appointees change polices by how these are interpreted (Lewis Reference Lewis2008), and this influenced BIN’s enforcement pivots. Under Roosevelt, Commerce & Labor Secretary Straus and Ellis Island Commissioner Watchorn were considered liberal on immigration, while the official between them, Commissioner-General of Immigration Frank Sargent, who headed BIN, was mainly interested in the distribution of immigrants. Under Taft, respectively, Nagel and Williams, who was a friend of the president, and Daniel Keefe were described as “restrictionists.” Even when there were deviations, such as Watchorn flipflopping and Nagel moderating Williams and Keefe’s “zeal,” on enforcement (Feys Reference Feys2013: 271, 285–88, 301–13; McKeown Reference McKeown2008: 254; Cannato Reference Cannato2010: 185–87, 228–29), the administrations’ postures held.
The significance of these appointees on enforcement, as my framework suggests, can be appreciated by comparing how they approached public-charge debarments. Otherwise, Watchorn and Williams’s tenures on Ellis Island are seen as relatively continuous (Feys Reference Feys2013: 302; Keeling Reference Keeling2012: 173–74; Pitkin Reference Pitkin1975: 42, 54, 65–66). In 1905, Watchorn eliminated the practice that arrivals carry $10; such a “monetary test” was not mandated (Tichenor Reference Tichenor2002: 135). For exclusion, in 1909, Williams set the bar at $25 (ARCGI 1909: 132–34). Shipping lines countered, by ensuring that passengers had access to this amount via family, charities, or migrant agents, which facilitated “the release of about forty percent” (Feys Reference Feys2013: 306–7). Nonetheless, those debarred as likely public charges mushroomed by 261 percent, from 4,401 to 15,907, between 1909 and 1910 (Fig. 1). The debarment rate increased from 1.37 to 2.28 percent and then decreased to 1.64 percent in 1913 (Table 1). This reduction is connected to Secretary Nagel instructing Williams against setting a monetary amount for entry in 1911 (Nagel Reference Nagel1911a; 1911b; Williams Reference Williams1911). In 1912, such debarments declined, from 12,004 to 8,152, and to 7,914 in 1913 (Fig. 1).
Regarding medical examinations, Roosevelt’s team eased them. Germ theory prompted physicians to “hunt for those carrying harmful pathogens on their bodies or belongings” (Kraut Reference Kraut and Ueda2006: 112–13). Secretary Straus challenged this practice by the Treasury Department’s Public Health & Marine Hospital Service (PHMHS). I reviewed the secretary’s diary and found that Straus told its Surgeon General, Walter Wyman, that certificates issued for “poor physique” were “stereotype[s].” Wyman concurred (Straus Reference Straus1907: 33–34). Consequently, between 1908 and 1909 the number of surgeon’s certificates for physical/mental defects decreased and the debarment rate of 1.37 percent was lower compared to the second period (Fig. 1, Table 1). Since BIN reported that conditions improved for immigration, these outputs were unlikely caused by the recession between 1907 and 1908 (ARCGI 1909: 3–4).
Even when PHMHS and BIN officials resisted liberal immigration policies, the secretary coaxed their execution. Straus “loosely” interpreted Section 19 of the Immigration Act of 1907 (Cannato Reference Cannato2010: 171–74), and since this gave the secretary discretion, it “opened the door for any diseased immigrant to appeal for treatment upon arrival” (Fairchild Reference Fairchild2003: 41). This allowed newcomers, who suffered from tuberculosis or loathsome/contagious diseases, to receive hospital care to recuperate and then gain admission. A review of the minutes of a meeting on Ellis Island in 1908 reveals that the secretary pressed physicians and commissioners, respectively, to be consistent on their diagnoses and to be amenable to hospitalization. Watchorn doubted the legality of “landing aliens for treatment” and Sargent opposed making this a “practice.” Both calculated that shipping companies would “immediately relax their vigilance” (“Report of Conference Held” 1908: n.p.–7, 11–15, 29–35).
From 1908 to 1909 debarments for loathsome/contagious diseases decreased by 18 percent, from 2,900 to 2,382 (Fig. 1). The recession affected these in 1908 since the number of migrant crossings from Europe declined by 66 percent for part of this year, from 1,357,000 to 456,000 (Keeling Reference Keeling2012: 217). But, with better economic conditions in 1909 and the debarment rate remaining at 1.37 percent, the secretary’s intervention seemingly contributed to the decline in these debarments.
The importance of political appointees for pivoting on enforcement is further seen in how Taft’s team strengthened medical inspections. With sharp increases in immigrant admissions in 1910 and 1913 (Fig. 1), and concerns about “inadequate” inspections by shipping lines, Nagel and Williams targeted newcomers with “mental or physical defect[s]” that might affect their capacity to work (ARCGI 1910: 134–35). Nagel asked Taft for more medical staff, especially psychiatrists, because it was “not adequate as to numbers” (Nagel Reference Nagel1912e: 2–4). This required more funding for PHMHS and BIN, which was required to reimburse the Treasury for medical screenings (US Congress 1907). The president fully supported enhancing these at Ellis Island (MacVeagh Reference MacVeagh1912a; 1912b; Taft Reference Taft1912).
These efforts apparently contributed to an increase in debarments based on surgeon’s certificates and loathsome/contagious diseases in 1913 (Fig. 1). Extra physicians expanded the “clinical gaze” used on inspection lines. Plus, advancements in medical technology “vastly improved the chances of diagnosing an excludable condition.” By 1910, X-ray technology was routine for diagnosing tuberculosis (Kraut Reference Kraut and Ueda2006: 113–14). However, these enhancements were offset by discontinuation of the monetary test since the debarment rate continued to decline between 1911 and 1913, from 2.48 to 1.64 percent (Table 1).
Nagel added to this reduction by rectifying Willliams’s harsh oversight of immigrant aid societies. These “tilt[ed] the scale in the immigrants’ favor when no one else would” (Cannato Reference Cannato2010: 215). Immigration regulations allowed societies to defend immigrants against exclusion and expulsion, while authorizing officials to verify the “good character and reputation” of societies’ staff (“Immigration Laws: Rules of” 1912: 43). Instead, Williams charged some organizations as “grossly mismanaged.” In 1909, the commissioner dismissed five, subsequently restoring three only after their “drastic reorganization” (ARCGI 1909: 134; Williams Reference Williams1909a; 1909b). In 1911, Nagel checked Williams by having rules developed on the expected character of societies (Nagel Reference Nagel1911b; Keefe Reference Keefe1912). In 1912, the numbers and rates of debarments and deportations decreased (Figs. 1–2, Table 1).
As my framework considers, centralization played a role in bolstering deportation via coordination with debarment. In 1903, Chinese inspection staff was transferred from the Treasury to the Commerce & Labor Department (Ettinger Reference Ettinger2009: 73–74). This integrated debarment and deportation functions. The Immigration & Chinese Division, one of BIN’s three units, was responsible for these functions (ARCGI 1906: 77–78; “Organization of the Bureau” 1907). Roosevelt and Taft’s appointees further refined the functions and focused on immigrants’ financial and health conditions. BIN’s centralized decision-making unevenly allowed appointees to manage two expulsion processes – for the Chinese and others – to return them to their countries.
Both administrations centralized deportation and debarment activities, consistent with the managerial presidency’s preference for functionality (Arnold Reference Arnold1998 [1986]: 20–21). This counteracted the Chinese entering via Canada and Mexico since major ports were reinforced (Hester Reference Hester2017: 21). Under Roosevelt, one supervisor was entrusted with the Southern line (ARCGI 1908: 161–62). Under Taft, supervision of both boundaries was solidified by consolidating staff performing immigration enforcement activities into 23 districts, because of “a lack of uniformity of methods” (ARCGI 1911: 167–71; “Immigration Laws: Rules of” 1912: 45).
These centralization efforts boosted functional specialization in this Division; staff were positioned for better horizontal coordination of these activities (Nou Reference Nou2015: 459). Personnel changed immigrants’ status from admitted to unlawful to expel them (Hester Reference Hester2017: 69), and the public charge category was “coupled” with others (Moloney Reference Moloney2012: 82). Staff homed in on public charges for physical/mental health, representing 35 percent of expulsions between 1907 and 1913 (Fig. 2). These buoyed the deportation rate, which was buttressed by the debarment rate. Because 76 percent of exclusions were for public charges and health reasons (Fig. 1). While proceduralism, which supported Chinese exclusion, is considered to have produced practices that influenced enforcement more than any appointee or bureau (McKeown Reference McKeown2008: 251–55), my analysis shows how organizational structure was just as important for BIN to hone enforcement.
Like remote control weakening debarment, global and domestic forces stymied deportation. American diplomats negotiated with their counterparts for the return of their citizens. Plus, BIN officials juggled two deportation protocols per Chinese Exclusion, and general immigration, laws. The Chinese used the former to go to federal courts to challenge expulsion efforts (Hester Reference Hester2017: 4–8, 16–23, 173–74). Officials worried about surreptitious entries (ARCGI 1906: 81). The numbers for arrests and deportations of the Chinese increased under both presidents, declining later in Taft’s term (Table 2).Footnote 3 Debarment and deportation rates for the Chinese generally followed the same pattern (Table 3).Footnote 4
Congress’s sway over BIN via budgeting was pronounced in the first period; increases to its annual funding supported enforcement (Fig. 3).Footnote 5 The Bureau regularly pleaded for “more money, more employees, and better facilities” (Schneider Reference Schneider2011: 75). It was federally funded through the head tax for admission, which grew from $0.50 to $4 between 1891 and 1907 (Ettinger Reference Ettinger2009: 70–71); generating $37 million from 1894 to 1913, while expenses totaled $29 million (ARCGI 1914: 32). Table 4 illustrates that, between 1907 and 1909, increases in total spending mostly resulted in increases in debarment and deportation rates. This differs from knowledge that enforcement is more responsive to decreases in funding (Wood and Waterman Reference Wood and Waterman1993), and that delays in responsiveness caused by organizational hierarchy (Carpenter Reference Carpenter1996) may well be reduced by centralization.
Figure 3. Expenditures (left scale) and employee positions (right scale).
Source: Own elaboration based on ARSCL, ARCGI, 1906–14 and RDCL, ARCN, 1913–14.
Note: Data for 1913 include estimates because of departmental reorganization.
Table 4. Trends in agency resources and immigration enforcement rates
Source: ARSCL, ARCGI, 1906–14 and RDCL, ARCN, 1913–14.
Between 1910 and 1913, Table 4 shows that increases and decreases in total spending substantially resulted in opposite effects: decreases and increases in enforcement. This is attributable to the moderation of Williams’s policies in 1911, when the debarment rate decreased and total expenditures increased in 1912. Also, Nagel regulated Keefe’s practices (McKeown Reference McKeown2008: 251–55), which may explain the decrease in the deportation rate in 1912 while total spending increased.
These effects show that Congress weakened its control of BIN; this observation was facilitated by my framework’s consideration of bureaucratic discretion over funding. In 1909, lawmakers terminated the Immigrant Fund (ARCGI 1909: 145), where head taxes up to $2.5 million were deposited (US Congress 1907), and which supported Bureau operations (Ettinger Reference Ettinger2009: 85). This termination eliminated BIN’s funding discretion since it would only receive congressional appropriations. This discretion may well have empowered agency officials over Congress and incentivized them to maximize their budget (Niskanen Reference Niskanen2008 [1971]). They merely refunded this tax when immigrants were deported (“Immigration Laws: Rules of” 1912: 21–22). Without this motivation, arguably, the agency became less responsive to lawmakers’ funding changes, and its personnel inclined to restrict immigration.
Additionally, Congress expanded BIN’s staff positions every year (Fig. 3), and these largely contributed to increases in debarment and deportation rates (Table 4). Except in 1907 and in 1912–13, respectively, because of Straus and Nagel’s aforementioned liberal decisions. Also, employee turnover may have contributed to these decreases; seaport and border stations were short-staffed (ARCGI 1914: 171–72, 198, 219–26, 246–50, 255). Nonetheless, the growth in positions aligned with the administrative state’s “almost five-fold” expansion of civilian employees between 1871 and 1901 (Arnold Reference Arnold2009: 12). They were indispensable, as immigration enforcement was labor-intensive. For exclusion, inspectors observed and questioned newcomers; in peak periods on Ellis Island, inspection time was two minutes (ARCGI 1909: 134; ARCGI 1914: 181–82). For expulsion, inspectors were supposed to “monitor immigrants, review documentation, detain and patrol, and coordinate” with governmental, medical, and charitable entities (Moloney Reference Moloney2012: 10).
The quality of staff was important for enforcement, too. Between 1907 and 1912, the Department of Commerce & Labor conducted efficiency ratings to determine how much and how well employees performed. Of 7,119, 14 percent were found to be unsatisfactory. This totaled 1,023; 35 percent were counseled, and 14 percent were demoted.Footnote 6 BIN used the ratings to transfer workers to facilities where their abilities were a better fit, and to train long-term employees who were inflexible (ARCGI 1911: 167–70). This aligned agency expectations with employee evaluation, assignment, and training. Plus, the Bureau relied on immigration regulations to limit inspectors’ discretion. Such standardization efforts make decision-making easier (Nou Reference Nou2015: 464), and predictable.
With respect to infrastructure, my framework helps to elucidate that it was slowly built, and BIN was forced to rely on temporary accommodations. Although funds were allocated for immigration stations (Fig. 3), their construction required lead time. The agency did not design and buildout facilities. Also, private and public entities provided makeshift spaces for immigration enforcement. These considerations detracted from the efficacy of stations, which were inadequate to meet inflows (Pitkin Reference Pitkin1975: 105–9; Schneider Reference Schneider2011: 74–79). Most construction was completed in the second period (ARCGI 1914: 34–35, 199), when debarment and deportation numbers and rates were greater than the first (Figs. 1–2, Tables 1, 3).
Infrastructure expenditures were primarily for Ellis and Angel islands. Enforcement lagged because the supervising architect in the Treasury Department managed construction projects, which were frequently delayed (ARCGI 1914: 34–35). Between 1907 and 1909, spending on immigration stations totaled $1.6 million; almost the full amount financed 17 new buildings and equipment for Ellis Island’s medical system, and the construction of Angel Island (ARSCL 1907: 64; ARCGI 1908: 135–36; ARSCL 1908: 88; ARCGI 1909: 145; ARSCL 1909: 72). From 1910 to 1913, this system supported health-related exclusions and expulsions. Taft directed the opening of Angel Island for adequate detention space on the West Coast (Nagel Reference Nagel1909). This sustained enforcement against the Chinese.
The immigration infrastructure – supported by federal funds – was compromised by remote control. BIN’s immediate needs were met by in-kind contributions from transportation and steamship companies and state and local authorities. Contributions from business interests constituted the “extractive border” and weakened enforcement (Feys Reference Feys2016: 78–79). Among others, the Bureau used the companies’ ships for inspection in Providence, and local jails run by sheriff offices and police agencies near the Northern border for detention (ARCGI 1913: 45; ARCGI 1914: 174). Such accommodations were not optimal since immigration officials neither designed nor controlled them. Immigration was supposed to be curtailed by naturalization enforcement, but as will be discussed next, it was not coordinated with immigration enforcement.
Tightening citizenship
BIN’s centralized structure helped political appointees to regulate naturalization, like with immigration. But, the Bureau’s “divisional form” concentrated coordination within its divisions (Nou Reference Nou2015: 459). The second of these, the Naturalization Division, was further reconfigured by Chief Richard Campbell to enhance enforcement. Also, Campbell standardized investigation of applicants and recommendations staff made to judges. Below I trace these efforts along with how the Division’s staff, budget, and infrastructure were adjusted to meet its “gatekeeper function” (Schneider Reference Schneider2001: 50). These resources were limited because the Division was inextricably tied to courts, where clerks processed naturalization petitions and judges decided them.
Great interest in naturalization was tempered by red tape. Fig. 4 shows that the sum of papers – declarations of intention, petitions, and certificates – multiplied by 255 percent from 101,221 to 358,835.Footnote 7 The overall denial rate of petitions increased from 3.1 to 16.6 percent between 1907 and 1910 and then decreased to 11.7 percent in 1913. This decline was furthered by financial support for court clerks. In Southern states, this rate mostly increased from 4.8 to 20.9 percent in the entire period, with marked growth since 1910. Because the Division paid attention to this region. Although not reported in 1907, for all denials the main reasons were want of prosecution – applicants gave up – and incompetent witnesses – they were deemed inadequate – accounting for 61 percent, 28,742 of 46,995 (Fig. 5).Footnote 8 These reasons illustrate how convoluted the process was (Gavit Reference Gavit1922: 126–29; Schneider Reference Schneider2011: 204–9).
As an appointee, Campbell was “firmly aligned with restrictionists” (Schneider Reference Schneider2011: 209), and in line with my framework, poised to interpret divisional policies in lockstep with them and superiors (Lewis Reference Lewis2008). The chief previously served on Roosevelt’s Naturalization Commission, which recommended that the Commerce & Labor Department execute naturalization laws (“Report to the President” 1905). Like immigration enforcement, Campbell reported to the same presidents, secretaries, and commissioners general. Both administrations agreed on the need to safeguard naturalization. Straus considered the old process “a disgrace” and Nagel thought the new one required “care” (Straus Reference Straus1908: 186; Nagel n.d.: 19).
For “screening out unsuitable petitioners” (Schneider Reference Schneider2011: 209), the chief fine-tuned administrative processes. In 1907, the Division’s clerks, at headquarters in Washington, DC, reviewed declarations and petitions. Although cursory, defects were detected (ARCGI 1907: 114–15; ARCGI 1908: 198). In line with standardization (Nou Reference Nou2015), this supported consistency between judicial decisions and naturalization laws. To bolster this, in 1910, Campbell adopted an investigative process, in-person and via mail; divisional examiners collected evidence on petitioners and witnesses for judges. For 1912 and 1913, summaries were presented at 95 percent of hearings, 11,537 of 12,125 (ARCGI 1911: 233; ARCGI 1913: 183; ARCN 1914: 23–24). The overall denial rate sharply increased in 1908 and gradually in 1909 and 1910, implying that these initiatives were effective (Fig. 4).
Also, my framework helps to understand that centralization seemingly affected these rates in contrasting ways. In 1910, Campbell reorganized the Division into 11 districts; nationally distributing work to offices in Boston, Chicago, Denver, New York, Philadelphia, Pittsburgh, San Francisco, Seattle, St. Louis, St. Paul, and Washington, DC. Each covered a region, except the Southern states, which continued to be neglected until 1911, when headquarters oversaw them (ARCGI 1907: 125; ARCGI 1910: 202–4; ARCGI 1911: 230–32). Such centralization presumably saved time for decision-making (Arrow Reference Arrow1974), expediting command of enforcement activities. With more attention on these states – Alabama, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, and Virginia – their denial rate increased, while the overall rate declined, and both dipped in 1913 (Fig. 4).
Congress increasingly funded the Division between 1907 and 1913 (Fig. 3), contributing to growth in the overall denial rate half the time (Table 5). Funds were collected via fees, up to $5 for filing a declaration and a petition (“Naturalization Laws and Regulations” 1913: 9–10), totaling $1.6 million (ARCN 1914: 35), and $1.1 million was used (Fig. 3). From 1910 to 1913, funding infusions apparently did not boost this rate because of asymmetry in the divisional budget for staff and court clerks (ARCGI 1910: 204, 226; ARCGI 1911: 240; ARCGI 1912: 197–98). Expenditures for field staff increased by 48 percent, from $108,607 to $160,495, while those for these clerks rose by 357 percent, from $8,599 to $39,264 (ARCN 1914: 22).
Table 5. Trends in naturalization process resources and denial rates
Source: ARSCL, ARCGI, 1906–14 and RDCL, ARCN, 1913–14.
This aid to court clerks, since 1910 (ibid.), counteracted naturalization restrictions. One study, often cited by scholars (Schneider Reference Schneider2001, Reference Schneider2011), considered divisional staff to be hyper-technical (Gavit Reference Gavit1922: 167–69). The Division reported that the decline in denials resulted from these clerks learning the Naturalization laws and paperwork (ARCGI 1912: 185–90). This knowledge likely allowed clerks to better assist applicants in the courts. In 1911, one of the noted reasons for denials, want of prosecution, decreased by 17 percent, from 3,697 to 3,070 (Fig. 5). For New York courts this decrease was greater, by 52 percent, from 1,512 to 721 (ARCGI 1910: 205, 213–14; ARCGI 1911: 220–22). These courts represented 22 percent of the caseload (ARCGI 1910: 220). In 1912, the “other” reasons, a catchall category, decreased by 25 percent, from 1,456 to 1,085 (Fig. 5). Unless expenses for interagency work were proportional, my framework facilitates learning that funding increases for divisional enforcement alone would not have resulted in its growth (Wood and Waterman Reference Wood and Waterman1993).
Also, Congress expanded the Division’s employee positions almost every year, contributing to the rise in the South’s denial rate (Table 5). Although staff growth generally facilitated the overall denial rate’s increases from 1907 to 1910 (Table 5), this rate then decreased as discussed. Presumably, the additional staff was insufficient to offset this rate’s decline because of disproportionate funding. Yet, the denial rate in the South increased precipitously in 1911 and 1912. This can be attributed to allotments of clerical funding to the busier courts in big cities (ARCGI 1909: 211–13; ARCGI 1910: 204–6, 219–20), leaving naturalization processing more dependent on divisional staff in this region. Plus, the Division benefited from personnel increases, including a major transfer of staff in 1910, from the Justice to the Commerce & Labor, Department (Table 6).Footnote 9 These increases complemented the Division’s subsequent dedication of an office for the South.
Campbell’s choices, including establishing processes to screen applications and to investigate petitioners, show how determinative political appointees can be over policy outputs (Calvert et al. Reference Calvert, McCubbins and Weingast1989: 604–5). The upward trend in the South’s denial rate implies that staff, who can deviate from protocol (Lipsky Reference Lipsky2010 [1980]), were mainly compliant. But it is unclear if the Department’s efficiency ratings program contributed to this, like with immigration enforcement.
Relative to my framework’s attention to infrastructure, the Division was better at influencing naturalization outside than inside BIN. In the US political system’s separation of powers, the judiciary granted citizenship. Campbell swayed judges through investigative summaries sent to them. They overruled his examiners’ objections in only 0.6 percent of cases, 877 of 155,258, in 1912 and 1913 (ARCGI 1913: 183; ARCN 1914: 23–24). This hints that naturalization laws were treated as “chips;” divisional staff and judges negotiated these objections. This constitutional constraint was mitigated, like the policy state does for policymaking (Orren and Skowronek Reference Orren and Skowronek2019: 6, 41), to buoy the declination rate.
The Bureau’s divisional form constrained intra-agency collaboration on infrastructure. While this form supports policy specialization, it can create inconsistencies (Nou Reference Nou2015: 460). This arrangement helped with coordination on naturalization restriction in the Division, but not with the Immigration & Chinese Division on diversifying locations for immigration stations. These were concentrated in the Northeast, as were almost half the naturalization cases (ARCGI 1907: 137–38; ARCGI 1910: 219–21). Taft encouraged immigration to other seaports by declining to invest in Ellis Island (Feys Reference Feys2013: 308). This could have changed this geographic concentration, which limited agency resources to one region and, by extension, opportunities for naturalization in others. Yet, Campbell’s annual reports do not reveal interest in capital projects for these stations. Next, I will explain the lobby pressures behind immigration and naturalization enforcement.
Coalition management
The market analogy referenced earlier illustrates BIN’s production of enforcement outputs as supply, and lobby pressures for and against immigration as demand. In this section, I will discuss these pressures on the Roosevelt and Taft administrations and their alternating enforcement responses (coalition management). The managerial presidency’s problem-solving capacity attracted interest groups to policy implementation (Arnold 1998 [1986]: 18–20; Reference Arnold2009: 9–13). In the Progressive Era, their activity is described as client politics since shipping companies most benefitted from open immigration (Feys Reference Feys2017: 368–69), while the public bore costs. From 1907 to 1913, this is better defined as interest group politics because opponents received concentrated benefits and costs (Freeman Reference Freeman1995), which fluctuated via these responses. This is further understood by considering that agency design can pit groups against each other (Macey Reference Macey1992), like the Immigration Restriction League and the Union of American Hebrew Congregations.
My framework helps to expose that BIN’s design was conducive to shift enforcement to respond to conflicting lobby pressures, resulting in a zero-sum game. These pressures were “fairly balanced” until the Great War (Marinari Reference Marinari2020: 15). BIN was structured like its parent Department, Commerce & Labor, to respond to lobbies interested in multiple-regulatory functions (Macey Reference Macey1992), on immigration and naturalization. This allowed lobbies to hold sway over policy formation (Lowi Reference Lowi1979). The Bureau responded by reducing debarment and deportation, and naturalization denial, rates to increase the numbers of immigrants and citizens. For capital owners, immigrant communities, and shipping companies (Goldin Reference Goldin, Goldin and Gary1994; Feys Reference Feys2017), this growth represented concentrated benefits since they enjoyed lower labor costs, more members, and higher revenue, respectively. For immigration opponents – labor unions and rural America, namely, the South and native-born (Goldin Reference Goldin, Goldin and Gary1994) – this increase reflected concentrated costs, including job competition and wage reductions. When BIN pivoted, so did these benefits and costs, promoting interest group politics since one side’s gain was another’s loss (Wilson Reference Wilson and James1980).
Concerning immigration enforcement, organized interests pressed for their policy preferences and political appointees answered. They helped their presidents develop policy solutions through administrative means (Arnold Reference Arnold2009: 66–70, 129–31), producing ebbs and flows, as previously explained, in the debarment rate via the monetary test. When Watchorn eliminated it in 1905, this aligned with immigration activists, like Jane Addams, who subsequently urged Congress not to specify a monetary amount in the public charge provision (US Congress 1906b: 8515). When Williams reimposed the test in 1909, the American Federation of Labor endorsed it as protective of US workers; later the Immigration Restriction League lobbied for legislation (“Statements and Recommendations Submitted” 1911, 41: 103, 369–72).
Competing lobby demands on immigrants’ medical conditions contributed to fluctuating exclusion and expulsion rates. Immigration activists secured inclusionary, among exclusionary, provisions in immigration laws to “push for liberalizing” (Marinari Reference Marinari2020: 15). When the National Liberal Immigration League contributed to a change in tone for fair treatment in immigration enforcement under Roosevelt (Cannato Reference Cannato2010: 174–75), Straus implemented such policies in medical care, as discussed supra. These rates were lower in the first, than in the second, period. Then chambers of commerce lobbied Nagel for more enforcement of the infirm (Giltner Reference Giltner1912; Gurley Reference Gurley1912; Plumb Reference Plumb1912). Also, the New York State Commission on Lunacy cited a significant decrease in certifications of the mentally ill under the secretary (Campbell Reference Campbell1912). With Taft’s support, Nagel and Williams’s previously discussed reinforcement of health and mental health screenings responded to these pleas.
Both sides deftly used the tactic of technical expertise, including procedure (Clemens Reference Clemens1997: 1–2, 186; McKeown Reference McKeown2008: 254–55), to press political appointees. In effect, they were cajoled to “live up to their own book of rules” (Alinsky Reference Alinsky1971: 128). For strict enforcement, a case in point is the Immigration Restriction League, which pressured Watchorn for better enforcement of the poor physique clause (Hall Reference Hall1908; Watchorn Reference Watchorn1908). This undermined Straus’s formerly noted liberal efforts on medical examinations; Watchorn and Sargent resisted the secretary’s discretion to provide hospital treatment to newcomers.
For the liberal cause, an example is the Union of American Hebrew Congregations, which questioned Nagel about boards of Special Inquiry. Because the secretary reversed them on exclusions half the time (“The Immigration Question with” 1911: 5). These tribunals reviewed debarment and deportation decisions by immigration staff (US Congress 1907). In line with my framework’s consideration of lobby pressures, I read 46 appeals that the secretary adjudicated between October 20, 1911 and February 19, 1913, from Nagel’s personal papers and found that in several instances, subordinates were overruled (Nagel Reference Nagel1912a, b, c, d). So, the Union aimed for the secretary to ensure that the boards followed agency rules.
For naturalization enforcement, immigration restrictionists and the Dillingham Commission expressed fears about assimilation, contributing to the naturalization denial rate mostly rising in the South between 1907 and 1913. They complemented Campbell’s aforementioned enforcement activities, particularly by devoting an office to this region. When Congress considered BIN’s creation, the Immigration Restriction League lobbied to limit immigration to the South because of such fears; those from Southeastern Europe were unwelcomed, while small families were acceptable (House Committee on Immigration and Naturalization 1906a: 151–54). The Commission, which released its study on immigration in 1911, legitimized these fears by concluding that arrivals, without families in America, were unassimilable. Their tendency was “not great” towards naturalization (“Abstracts of Reports of” 1911, 1: 42).
Immigration activists counteracted Campbell by helping immigrants with naturalization, facilitating a decline in the overall denial rate since 1910. Americanizers offered them English and US history courses (Schneider Reference Schneider2011: 153–56). These likely mitigated petitioners from giving up, an aforementioned reason for denials, because they were more knowledgeable about applying for citizenship. Although naturalization laws mitigated corrupt practices to enlarge voter rolls (Schneider Reference Schneider2001: 55–57), political parties assisted them (Buenker Reference Buenker1988: 195). Parties surely facilitated relationships between Americans and applicants to alleviate the disqualification of their witnesses, another reported reason for denials. Consequently, Southern and Eastern European immigrants applied for citizenship. In 1914, 68 percent of 26,284 petitioners hailed from Russia, Italy, Austria, and Hungary (Gavit Reference Gavit1922: 229).
As my framework helps to reveal, from 1907 to 1913 BIN’s enforcement pivots contributed to the equilibrium between opposition groups. This aligns with the view that Americans wanted neither a fully opened, nor fully closed immigration system (Cannato Reference Cannato2010: 10–11). The symmetry of these opponents illustrates interest group, not client, politics. The mode of immigration politics can change through “particular immigration cycles” (Freeman Reference Freeman1995: 886–87). This is better appreciated by considering naturalization enforcement since shipping companies’ efforts against restrictionists “overlapped” with the “immigrant vote,” which was a “weightier factor” (Keeling Reference Keeling2012: 146–47). Urban immigrant communities swayed their representatives in Congress, including as they “gained the vote,” to oppose restrictive measures (Goldin Reference Goldin, Goldin and Gary1994: 225, 255). However, the structural arrangement that facilitated this equilibrium was significantly changed when these functions were separated, as I will examine next.
System redesign
In 1913, the Bureau ceased equally responding to opposing interest groups. In this section, I will explain why Congress divided BIN into two bureaus, resulting from its adverse social, economic, and political aftereffects on immigrants and, by extension, cities (system redesign). External forces played a visible role in this change; the anti-immigration coalition pushed lawmakers. The Dillingham Commission helped them by identifying these aftereffects and recommending immigration restrictions for economic reasons (Zeidel Reference Zeidel2004: 3–5, 145). Less visible internal forces prompted this reorganization, too. The Bureau’s enforcement pivots contributed to the institutionalization of immigrants in society and to their naturalization in the Northeast. These pivots represented the institutional change process of conversion; policies were repurposed for new aims (Hacker et al. Reference Hacker, Pierson, Thelen, Mahoney and Thelen2015: 185), and they produced other outcomes (Mahoney and Thelen Reference Mahoney, Thelen, Mahoney and Thelen2009: 2–4). BIN’s Division of Information was supposed to mitigate these aftereffects by encouraging newcomers to move from the Northeast but was under-resourced.
Created by the Immigration Act of 1907 as the Bureau’s third unit, the Division was unable to significantly reduce the concentration of immigrants in urban centers. Between 1908 and 1913, 121,477 individuals requested employment information from the Division and 27,398 or 23 percent, including US citizens, relocated among the states.Footnote 10 Meanwhile, there were 5.5 million arrivals (Table 1). Division Chief Terence Powderly was proficient on European emigration (Powderly Reference Powderly, Harry, David and Paul1940: ix–x, 302–6), and adopted liberal immigration policies (Feys Reference Feys2013: 291–92), under Roosevelt and Taft. But, funding constrained staffing and infrastructure (Leonard Reference Leonard1980: 211–12). The Division was allocated annually $19,340 in federal funds for nine administrative personnel in its Washington, DC, headquarters, and for the rent of two offices in New York and Chicago. Three others were at immigration stations in Baltimore, Galveston, and New Orleans. Immigration inspectors also doubled as divisional staff (ARCGI 1908: 173–82; ARCGI 1909: 235–36; ARCGI 1913: 218; Powderly n.d.).
Moreover, the Commission’s data suggest that BIN’s targeting of newer immigrants for medical reasons – discussed above – resulted in their overrepresentation in hospitals. In New York City, Commission agents visited hospitals to talk with patients between August 1, 1908 and February 28, 1909. Of 23,758, 52.3 percent or 12,426 were foreign-born; 38.1 percent of the foreign-born lived in America for 20 or more years, and 17.9 percent for under three years. Most long-term residents were of the older immigration; but, the newer immigration – namely, Jews and Italians – constituted 29.2 and 31.2 percent, respectively, of short-term residents. For physical health, 11.2 percent of the foreign-born were diagnosed with epidemic diseases and tuberculosis. For Jews and Italians, 12.8 and 15.6 percent, respectively, were treated for these. For mental health, 7.6 percent of the foreign-born were diagnosed with insanity; for this, these groups represented 10.4 and 7.3 percent, respectively (“Abstracts of Reports of” 1911, 2: 257–58, 264–67, 277–78, 290).
Also, the Commission’s study hints that newcomers were more likely to need charitable assistance since the Bureau’s previously mentioned enforcement interrupted their connections to immigrant-aid societies on Ellis Island. Between December 1, 1908 and May 31, 1909, the Commission designed a survey and trained most of the charities that administered it to families in 43 cities nationwide. Of 119,028 people who received assistance, 50,312 or 42.3 percent were foreign-born. Of all people, 61.8 percent needed assistance with unemployment/insufficient earnings and 21.8 percent received help to secure a job; of these, 23.2 percent were foreign-born. The numbers of older and newer immigrants were almost even in benefitting from any assistance, except that more of the foreign-born, 33.9 percent, resided in the states 20 or more years compared to 5.5 percent for three years. Among these short-term residents, more were from Southern and Eastern Europe: Italians represented 9.6 percent and Jews 9 percent (“Immigrants as Charity Seekers” 1911, 34: 3–4, 14–15, 40, 48–49, 73–74).
Both these studies shed light on the intractability of distribution. Their shortcoming is that they offer a snapshot for 1909 of the shift in immigration enforcement between the end of the Roosevelt administration and the start of the Taft administration. Still, as my framework suggests, I considered BIN’s impact beyond its enforcement outputs, and the studies’ data imply that the Bureau exacerbated resettlement for newcomers. The Division focused on farms as a destination, which were isolated for immigrants “accustomed to … omnipresent communal ties.” Plus, they were “desperately poor,” so transportation and relocation costs were prohibitive (Leonard Reference Leonard1980: 212–15). Put differently, these two considerations – a weak personal network in case of unemployment and a questionable return on investment for such costs – mirror the risk calculus migrants use to decide whether to travel to America (Keeling Reference Keeling2012: xiii–xiv, 38–39).
The Commission’s data further imply that naturalization enforcement was socially detrimental to children of newcomers. This enforcement concentrated these families in the Northeast, consistent with the overall naturalization denial rate decreasing and this rate increasing for the South, as explained before. Also, such trends discouraged their distribution there; from 1908 to 1913 the Information Division reported that 2,173 people, or 8 percent, moved to this region.Footnote 11 One aftereffect of BIN’s activities was juvenile delinquency. Based largely on available statistics, the Commission concluded that delinquency was prevalent in North Atlantic states because of immigration. But, in the New York children’s courts for the calendar year 1908, 11,523 or 85 percent of convicted minors were US citizens; only 2,085 or 15 percent were foreign-born. Of this group, there was a propensity to confine Italians and Russians, who constituted 33 and 37 percent, respectively (“Abstracts of Reports of” 1911, 2: 163; “Immigration and Crime” 1911, 36: 9–10, 310–11).
The incarceration of these children surely produced political aftereffects on their families. While citizenship was attractive for such benefits as union membership and federal and municipal employment (Schneider Reference Schneider2011: 204), naturalization was improbable for families separated this way. Because the Commission asserted that children were “the most potent influence in promoting the assimilation … through contact with American life in the schools … of their parents” (“Abstracts of Reports of” 1911, 1: 42).
Like castling in chess, immigration restrictionists secured reorganization of the immigration system to protect and advance the policies that followed. As interest groups that formerly favored immigration joined this coalition (Goldin Reference Goldin, Goldin and Gary1994: 255), in 1913, the American Federation of Labor procured a new Labor Department to manage immigration (House Committee on Labor 1912). Congress also established bureaus for Immigration and Naturalization therein (US Congress 1913). This coalition structurally tilted bureaucracy towards labor by decentralizing these functions, which weakens the political control of elected officials (Whitford Reference Whitford2002: 186–87), who might ease enforcement. With events like World War I, enactment of the literacy test, and passage of the Nationality Quota laws approaching (Feys Reference Feys2013; Goldin Reference Goldin, Goldin and Gary1994; Keeling Reference Keeling2012; Marinari Reference Marinari2020), the immigration system was readied for closure.
Conclusion
Consistent with my core argument, this fine-grained narrative shows that Republican officeholders used a newly designed bureau to pivot between easing and tightening enforcement. Like an analytical framework is supposed to do, the three steps I discussed from my framework identified factors affecting policy implementation and their relationships (Schlager Reference Schlager and Sabatier2007: 293). Below I will explain the insights I developed from evidence previously considered in immigration history literature, and how these add to this, and political control, scholarship. The Roosevelt and Taft administrations responded to conflicting lobby pressures, but failed to coordinate BIN’s immigration, naturalization, and distribution functions. Public agencies are not designed for effectiveness (Moe Reference Moe, John and Paul1989), and this failure harmed immigrant communities, which, paradoxically, were blamed for urban social, economic, and political ills.
The resource adjustment step helped to elucidate several strengths of political control of bureaucracy. In line with scholars in this literature, Roosevelt and Taft used BIN to pivot on enforcement, and to facilitate immigration distribution, via centralization and political appointments. Centralization permitted quick execution since it reduces hierarchical layers in decision-making (Arrow Reference Arrow1974; Nou Reference Nou2015) and proved as influential as proceduralism – a driver of immigration enforcement (McKeown Reference McKeown2008) – to effectuate these shifts. Political appointees, too, were significant since they largely implemented policy decisions aligned with their presidents (Lewis Reference Lewis2008; Moe Reference Moe, John and Paul1989). These findings are in accord with immigration history scholars who have examined these presidents’ appointees and their decisions in immigration bureaucracy (Feys Reference Feys2013; Keeling Reference Keeling2012). My analysis contributes to political control literature by illustrating that such pivots need not result from the tension between the president and Congress (Whitford Reference Whitford2005); these can be produced by their relative cooperation.
There were weaknesses in the political control of BIN. Additional staff generally facilitated enforcement, yet funding schemes proved consequential. With immigration, strict enforcement under Taft was furthered by officials who lost discretion over funding (Niskanen Reference Niskanen2008 [1971]), disincentivizing them from admitting newcomers to collect head tax. This insight is new, even for immigration history scholars who have examined the Bureau under this president (Feys Reference Feys2013; Keeling Reference Keeling2012). With naturalization, enforcement was more equitable for petitioners when funding for court clerks was higher than for BIN staff; the overall denial rate declined since 1910. This nuance in funding is important for political control literature since changes to funding levels not only affect agency operations (Wood and Waterman Reference Wood and Waterman1993) but also interagency efforts.
Another flaw in political control was the Bureau’s infrastructure, having a domino effect on its three divisions. Immigration facilities, concentrated in the Northeast, were overused and induced arrivals to settle there. This pattern was reinforced by higher naturalization denial rates in the South and lower rates overall, including New York, further discouraging immigrant relocation. These insights are distinct from immigration history scholars who see the facilities as weak on enforcement (Pitkin Reference Pitkin1975; Schneider Reference Schneider2011), but not diminishing the divisions. BIN failed to mitigate this effect; its divisional form limited coordination to each unit (Nou Reference Nou2015).
The coalition management step helped to expose the mode of immigration politics as interest group politics from 1907 to 1913. Instead, immigration history literature identifies the mode as client politics in the Progressive Era (Feys Reference Feys2013). This change foreshadowed that immigration bureaucracy would transition from open to closed (Freeman Reference Freeman1995). Yet, the zero-sum game between coalitions for and against immigration is not apparent without considering BIN’s enforcement pivots. These have been taken for granted in this literature since debarment and deportation rates only gradually increased (Pitkin Reference Pitkin1975: 42–44, 98). Also, the literature focuses more on Chinese exclusion to understand how lobby groups used inclusionary and exclusionary provisions of immigration laws to influence enforcement (Marinari Reference Marinari2020: 186).
The system redesign step facilitated a better understanding of BIN’s reorganization in 1913. Structural politics makes clear how the Bureau and Commerce & Labor Department were reconfigured by Congress, as immigration restrictionists demanded (Moe Reference Moe, John and Paul1989). But the process of conversion, when rules are reinterpreted and aftereffects produced (Hacker et al. Reference Hacker, Pierson, Thelen, Mahoney and Thelen2015; Mahoney and Thelen Reference Mahoney, Thelen, Mahoney and Thelen2009), illustrates that BIN contributed to its reorganization. I traced the Bureau’s enforcement outputs, using the Dillingham Commission’s study, to social, economic, and political problems in immigrant communities. Otherwise, these issues were attributed to newcomers. This insight into institutional change rooted in bureaucracy benefits immigration history and political control literature by expanding scrutiny from external to internal forces.
Lastly, the lesson my narrative offers is that agency design, in the Progressive Era’s nascent administrative state, proved a double-edged sword: providing leeway in policy implementation while exacting a toll on immigrant communities. This did not require investigation of a lengthy period of time either, which institutional change literature otherwise suggests (Mahoney and Thelen Reference Mahoney, Thelen, Mahoney and Thelen2009: 2–3). Plus, my framework is portable (Pierson Reference Pierson2004: 6), and can be used to study other periods of immigration enforcement. This can further the literature discussed herein. Because sharing insights on bureaucracy across disciplines (Hacker et al. Reference Hacker, Pierson, Thelen, Mahoney and Thelen2015: 204), can improve understanding of the immigration state.
Acknowledgements
I thank the anonymous reviewers for their comprehensive comments and the editorial team for their support. Also, I acknowledge the City University of New York Faculty Fellowship Publication Program, including Dr. Lina Newton, for help with an earlier version of the manuscript. I further credit my colleagues, Drs. Jonathan Engel and Jerry Mitchell, for their helpful suggestions on more recent versions of the manuscript. It should also be noted that the Marxe Dean’s Research Award 2024–2025 supported making this article open access. Finally, I dedicate this research to the memory of my dear sister, Soraya Hernandez, who succumbed to her battle with cancer in 2024. She will always be my inspiration.
Open access
